You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am trying within our org due to existing vpc restrictions.Appreciate that you've made this available now to existing vpc.
As a result we've created our existing eks cluster as well.
Our VPC's private subnets are non routable to the internet directly(either via NATGW or IGW).
We use http squid proxy for outbound connectivity.On this subnet we created managed nodegroups
with http proxy environment settings.
Based on the above scernario how do accomplish enabling litellm via http proxy for outbound
"If you'd like to run LiteLLM in subnets without outbound internet
access, set DISABLE_OUTBOUND_NETWORK_ACCESS="true". Due to lack of
internet access, this configuration will only work with AWS Bedrock or
SageMaker models. Because of this, you must remove all
non-Bedrock/non-SageMaker models from your config/config.yaml file. If
you do not do this, LiteLLM will fail to start as it will attempt to
call third party models over the internet."
Thanks
The text was updated successfully, but these errors were encountered:
Hi @mirodrr2
I am trying within our org due to existing vpc restrictions.Appreciate that you've made this available now to existing vpc.
As a result we've created our existing eks cluster as well.
Our VPC's private subnets are non routable to the internet directly(either via NATGW or IGW).
We use http squid proxy for outbound connectivity.On this subnet we created managed nodegroups
with http proxy environment settings.
Based on the above scernario how do accomplish enabling litellm via http proxy for outbound
"If you'd like to run LiteLLM in subnets without outbound internet
access, set DISABLE_OUTBOUND_NETWORK_ACCESS="true". Due to lack of
internet access, this configuration will only work with AWS Bedrock or
SageMaker models. Because of this, you must remove all
non-Bedrock/non-SageMaker models from your config/config.yaml file. If
you do not do this, LiteLLM will fail to start as it will attempt to
call third party models over the internet."
Thanks
The text was updated successfully, but these errors were encountered: